checkpoint @ game=820 steps=820

training_log.txt

@@ -746,3 +746,4 @@
 [2026-06-12 03:13:56] game=  817 | 1/2-1/2 | moves= 51 | buf=50,000 | steps=817 | loss=nan (p=nan v=0.5917) | game=10.8s train=162.4s
 [2026-06-12 03:16:54] game=  818 | 1/2-1/2 | moves= 53 | buf=50,000 | steps=818 | loss=nan (p=nan v=0.5447) | game=11.5s train=165.9s
 [2026-06-12 03:19:54] game=  819 | 1/2-1/2 | moves= 59 | buf=50,000 | steps=819 | loss=nan (p=nan v=0.5996) | game=13.1s train=166.5s
+[2026-06-13 03:58:58] game=  820 | 1/2-1/2 | moves= 81 | buf=50,000 | steps=820 | loss=nan (p=nan v=0.5778) | game=23.9s train=18.0s